The implementation of formal root cause analysis was associated with substantial improvements in most measured patient outcomes on acute pain services at three hospitals. The rates of overall events, respiratory depression, severe hypotension, and pump programming errors decreased, but incidence of severe pain increased.